You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tinkerpop.apache.org by "stephen mallette (JIRA)" <ji...@apache.org> on 2018/07/12 14:54:00 UTC
[jira] [Closed] (TINKERPOP-1999) [Java][gremlin-driver] Query to a
remote server via the websocket client hangs indefinitely if the server
becomes unavailable
[ https://issues.apache.org/jira/browse/TINKERPOP-1999?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
stephen mallette closed TINKERPOP-1999.
---------------------------------------
Resolution: Fixed
Assignee: stephen mallette
Fix Version/s: 3.2.10
3.3.4
3.4.0
Thanks for reporting this - I think i've fixed this. Some description of the fix can be found on the commit comments here:
https://github.com/apache/tinkerpop/commit/3dcabd4f1b23fecde28c79a10502ad562d934c8d
> [Java][gremlin-driver] Query to a remote server via the websocket client hangs indefinitely if the server becomes unavailable
> -----------------------------------------------------------------------------------------------------------------------------
>
> Key: TINKERPOP-1999
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1999
> Project: TinkerPop
> Issue Type: Bug
> Components: driver
> Affects Versions: 3.3.2, 3.3.3
> Reporter: Ankesh Khandelwal
> Assignee: stephen mallette
> Priority: Minor
> Fix For: 3.4.0, 3.3.4, 3.2.10
>
>
> Query to a remote server via the websocket client hangs indefinitely if the server becomes unavailable.
>
> Consider the code snippet below.
>
>
> {code:java}
> // Create a websocket client
> final Client client = Cluster.build().addContactPoint(<endpoint>).create().connect();
> // submit a query
> ResultSet resultSet = client.submit("g.V().id()");
> // wait for all the results
> resultSet.all().join();
>
> {code}
>
> If the
> {code:java}
> client.submit(...){code}
> call succeeds (that is the server is available/reachable) but before getting the results from the server, the server becomes unavailable/unreachable, then the client code hangs indefinitely at
> {code:java}
> resultSet.all().join(){code}
> .
> The right behavior would be to throw an exception indicating the server is not available anymore. The hanging behavior suggests that the *channelInactive* call from websocket connection may not be getting propagated properly (perhaps getting swallowed somewhere).
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)